Here's a familiar sort of story — a party loses an international arbitration, loses again when it tries to overturn the award at the seat, and then, instead of paying up, waits for the winner to come collecting in another country, and re-runs the exact same arguments all over again, this time dressed in the local law's clothing. Different courtroom, different flag, same losing case. And for years, that second bite was worth taking, because it so often worked.
In Nagaraj V. Mylandla v. PI Opportunities Fund-I, the Supreme Court has, for the first time, decided to do something about it. The tool it reached for has an intimidating name — "transnational issue estoppel" — but the idea underneath is refreshingly plain: if a competent court at the seat of the arbitration has already decided an issue between the same two parties, an Indian court asked to enforce the award will not let the loser fight that same issue all over again. Or, as the Court itself put it rather more vividly, a foreign award-debtor gets only "one bite at the cherry."
It is a quiet judgment about a noisy, expensive problem. Let's walk through it.
The Dispute, in Brief
The facts will feel familiar to anyone who has been near a venture deal gone sour. A 2014 shareholders agreement tied together a company, its promoters (the Mylandlas), and a set of investors holding a 51% stake. The investors were promised a way out — principally a Qualified IPO by 2016, with some fallback exit rights if the IPO never happened. The IPO never happened. The exit never came. And the whole thing curdled into a SIAC arbitration seated in Singapore, which, in 2024, the investors won — walking away with a hefty damages award reported to be somewhere near USD 121 million.
The Mylandlas did what losing parties tend to do — they went to the Singapore High Court, the court supervising the seat, and asked it to set the award aside, arguing the tribunal had ignored a "waiver" defence and a "buy-back" defence. The seat court was unpersuaded and upheld the award in 2025. Now here's the detail that ends up mattering enormously: the Mylandlas did not appeal any further to the Singapore Court of Appeal. They let it rest.
So the investors came to India to collect, filing to enforce the award before the Madras High Court under Part II of the Arbitration and Conciliation Act, 1996. And the Mylandlas reached straight back into the same bag — the award offended Indian public policy, the alleged share buy-back breached the Companies Act, 2013, there were complaints under the Specific Relief Act, election of remedies, natural justice — the entire orchestra, playing the same tune. The Madras High Court declined to conduct the concert. Enforcement, it held, is not a fresh trial on the merits; these issues had already been settled by the Singapore court; and transnational issue estoppel barred the encore. And so, up to the Supreme Court it went.
What the Court Actually Held
The headline is simple: the Supreme Court recognised transnational issue estoppel for the first time, openly noting there was no earlier decision of the Court on the point. Once you clear away the Latin, the principle is common sense wearing a robe — where a competent foreign court at the seat has finally and conclusively decided an issue between the same parties, an Indian enforcement court will not reopen it merely because the fight has been moved to a new building in a new jurisdiction.
But bare recognition would have been a hammer where a scalpel was wanted. The truly useful part of the judgment — the bit worth clipping and keeping — is a distinction the Court borrowed from Singapore, between two kinds of issue.
A forum-connected issue is one the enforcement court has to decide for itself, because it depends on the enforcing country's own law, public policy, or mandatory standards. India's fundamental policy is the obvious example. These are not estopped — they belong to the Indian court, and stay with it.
A forum-neutral issue is everything else the seat has already worked through — whether the agreed procedure was followed, whether the tribunal overstepped its jurisdiction, and, most often, the factual and contractual findings sitting underneath the award. These are the natural home of transnational issue estoppel. Once the seat court has decided them, the Indian court does not get a second go.
And this is exactly where the Mylandlas came unstuck — which is what makes the case such a useful teacher. Their clever move was to take a settled fact and re-label it as an illegality. The seat court had found, in substance, that there was no buy-back at all — the investors had simply agreed to hand back their shares in exchange for the money. The Mylandlas tried to recast that very finding as a Companies Act violation, and therefore a breach of Indian public policy. The Supreme Court was not fooled. Public policy, it held, is not a trapdoor back into the merits. And here is the sharp edge worth remembering: a public-policy challenge survives only where the seat court accepted the ground but withheld relief, not where the seat court looked at it and threw it out. Lose the point at the seat, and you have spent your bite.
So Has India Signed a Blank Cheque to Singapore?
No — and the Court took some care to say so. Along with the doctrine, it imported the brakes that come attached to it in the jurisprudence it drew from. Transnational issue estoppel will not apply to questions the enforcement court must answer under its own law; where the foreign judgment is territorially limited; where fairness calls for caution because the party now resisting estoppel never chose the foreign forum to begin with; or where honouring the foreign decision would collide with the enforcing court's own public policy. This is deference with conditions attached, not deference with the eyes closed.
That third brake — the one about who chose the forum — is worth tucking away. The classic ingredients of issue estoppel are a final judgment, from a competent foreign court, on the merits, with the same parties and the same subject matter. None of that asks whether you wanted to be in that court. The fairness exception does, and it leans in favour of a respondent hauled to a seat it never picked, over a claimant who chose it and now has to live with it.
What This Means if Your Disputes Cross Borders
Here is where the doctrine stops being an academic curiosity and starts rearranging the furniture.
Firstly, the seat fight is now the main fight.
If you plan to resist an award, the set-aside application at the seat is, in practice, your one real shot at the forum-neutral issues. You throw everything you have at it. The Mylandlas' unforced error — not carrying their challenge up to the Singapore Court of Appeal — is the cautionary tale wearing a sign around its neck. The plan of "we'll save that argument for the enforcement court in India" is now, more or less, a plan to lose slowly and expensively.
Secondly, it is use-it-or-lose-it — even for points you did not run.
The Court treated objections that had been raised, or could have been raised, before the seat court as foreclosed. Holding a forum-neutral argument in reserve for India may simply mean you have thrown it away. Put it all on the table at the seat.
Thirdly, sort your objections into two piles early.
Your genuinely Indian grounds — the fundamental policy of Indian law and its cousins — survive as forum-connected issues. Plead them, preserve them, lead with them. Do not burn your Indian hearing re-litigating merits the seat has already closed, because that is precisely the move this judgment was written to stop.
And a fourth point that is really the whole game — identity of issues is the battleground that is left.
Estoppel only bites where the Indian court is being asked the same question the seat answered. Where two courts apply two different legal systems to the same practical question, the "identity of issues" requirement tends to fail, and no estoppel arises — the point the UK Supreme Court made in Kabab-Ji, on whether a party was even bound by the arbitration agreement in the first place. So the shrewd respondent will work hard to frame its Indian objection as a genuinely different question of Indian law, and not a repainted version of what the seat already decided. Expect that characterisation contest — is this really a new question, or the old one in disguise? — to be where these cases are now won and lost.
One more thing, and it cuts both ways: this is not a one-way ratchet that only ever rewards the winner. If you are the award creditor but you happened to lose some sub-issue at the seat, that finding can bind you in India too.
There is a drafting lesson lurking in all of this as well. Because the seat court's rulings now travel with the award into Indian enforcement, the seat you casually agree to in an arbitration clause carries real weight down the line. Pick a respected, supervision-serious seat, and you strengthen the finality of any award you may one day need to enforce. Pick one carelessly, and you may be quietly locking in findings you cannot escape later in India.
Two Honest Caveats
First, this is one judgment, not a settled canon. The commentary has already spotted the loose thread — the Court applied the doctrine but did not quite decide whether transnational issue estoppel is a free-standing principle of common law, or simply a judge-made gloss on the discretion already living inside Section 48. That question will be litigated. Treat Mylandla as a door swinging open, not a matter closed.
Second, the case is about foreign arbitral awards, specifically. The similar estoppel that can flow from foreign court judgments travels a different road — Section 13 of the Code of Civil Procedure — and the two are not the same animal. If your cross-border headache is ordinary litigation rather than arbitration, the logic rhymes, but do not assume the framework transplants cleanly.
The Takeaway
For years, hopeful award-debtors treated the Indian enforcement court as a sort of appellate annexe to the seat — a place to run the merits one more time, under a public-policy flag of convenience. Mylandla is the Supreme Court, politely but firmly, declining the role. One bite at the cherry — and once you have had it, the question is closed.
If you enforce foreign awards, this is now your first-choice authority for shutting down the recycled objection. If you resist them, the honest ground left to you is genuinely forum-connected public policy — not merits in fancy dress. Either way, the centre of gravity has shifted back to where it probably always belonged: the seat.